GPU Cloud
L4, dedicated.
A single NVIDIA L4 GPU, 24 GB GDDR6 on the Ada Lovelace architecture, on an x86 host. $0.49 per hour, billed per minute, the same terms as every other GPUwerk tier.
Rented first come, first served: deploy an L4 when capacity is free, or join its waitlist in the console.
Specifications
Per GPU. The GPU and its memory are yours alone for the rental.
| GPU | L4, Ada Lovelace architecture |
|---|---|
| Memory | 24 GB GDDR6 |
| Memory bandwidth | 300 GB/s |
| CUDA cores | 7,424 |
| Max power | Up to 72 W |
| Host | x86 host |
| Tenancy | The GPU is yours alone while you rent it, never shared |
GPU figures from NVIDIA L4. GPUwerk publishes tok/s and training throughput figures only for runs on its own hardware. See the DGX Spark fleet benchmarks for how they are reported.
What would you run on it?
A low-power inference card: 24 GB at 72 W, for models that fit and traffic that does not need an H100.
Small-model inference
Chat or completion models up to about 14B parameters at FP8, and embedding or reranking models behind an API, where 24 GB is enough.
Video and vision pipelines
Hardware video encode and decode next to the GPU, for transcoding or for vision models that read video frames.
Low-cost development
Building and testing a serving stack on a cheap GPU before moving it to a larger one.
Poor fit: training and large fine-tunes
24 GB leaves little room for gradients and optimizer state, so anything beyond a small LoRA belongs on an H100 or larger.
What fits in 24 GB
Weights only: memory divided by bytes per parameter. Leave headroom for the KV cache, activations and the runtime, often 10 to 30% more.
| Precision | Per parameter | Largest model, weights only |
|---|---|---|
| FP16 / BF16 | 2 bytes | about 12B parameters |
| FP8 | 1 byte | about 24B parameters |
| 4-bit quantized | 0.5 bytes | about 48B parameters (no native FP4 on Ada; the runtime dequantizes) |
Billing
| Rate | $0.49 per hour, billed per minute |
|---|---|
| Billing | Per minute, against a prepaid credit balance |
| Fees | No egress fees |
| Commitment | None |
Prices are in USD, excluding applicable tax. Full billing mechanics, including what a stop versus a terminate does, are on the pricing page.
FAQ
Can I rent an L4 now?
Deploy it from the console when capacity is free; while none is, the console offers a waitlist and people in line are emailed one at a time, oldest first. Capacity is not held for anyone: whoever deploys first gets it.
How does billing work on this tier?
The same as every GPUwerk tier: per minute, against a prepaid credit balance, in USD, excluding applicable tax, with no egress fees and no commitment.
Need GPU time today?
A DGX Spark has 128 GB of unified memory and is rentable now, at $0.79/hour billed per minute. First top-up: pay $20, get $30 in credit. Or deploy an L4.